Earthquakes in the northern hemisphere during the years 1964-1970 were examined for a sidereal period. High chi square values were found but were lowered by canceling the 'midnight peak' in the network detection capability. The 1964 Alaskan earthquake and the 1965 Aleutian earthquake produced large numbers of aftershocks, which are believed to be the main cause of the bias which produced the high chi square values.

Earthquake aftershock sequences for the years 1964 through 1974 were examined for a sidereal period. High chi‐square values were found for various subsets of the data but were shown to be caused primarily by one specific sequence. Removing part of that sequence from the data eliminated the high chi‐square values when weight factors which eliminate local time fluctuations were used.

A reliable, inexpensive method for sending limited daily seismic data messages from remote observatories to the National Earthquake Information Center has been developed for use with the Argos satellite system. Data messages are compressed on a microcomputer and passed automatically to a simple transmitter. About 4 hr later, the data are available at the National Earthquake Information Center, where they are decompressed and reformatted into standard telegrams for use in quick epicenter determinations. Epicenter data are available daily to the international scientific community.

In a recent paper (Bune et al. 1970), a comparison of USSR and NOS (formerly C & GS) values was made and suggestions given for modifying NOS procedures. The paper also pointed out some of the problems of magnitude determination. We would like to present some of our data and make appropriate comments.Computation of body wave (mb) magnitudes by NOS was initiated in February 1963 with the use of Gutenberg & Richter's (1956) formula: A = ground amplitude in microns, recorded on short-period, vertical component instruments, T = period in seconds, and Q = distance/depth calibration factor. Since then, the NOS has computed over 25000 magnitudes. Contrast this with Gutenberg & Richter (1954) who used less than 1000 earthquakes to cover worldwide seismicity.It is the intention of this communication to examine items raised in the Bune, et al. (1970) paper and other more relevant aspects of magnitude determination. As a first step in this direction, we present a comparison of the USSR network data taken from trimonthly bulletins and the NOS data taken from PDE reports.To determine the relative capability of the two networks, a simple count of located events was made for the June to August 1968 period. The total of 1798 events consists of 1390 located by the NOS only, 383 located by both networks, and 25 located by USSR only. The NOS located over four times as many events as the USSR (1773, NOS; 408, USSR), a result which is to be expected based on the relative size and distribution of stations in the two networks. A comparison of magnitudes determined by the two systems is somewhat less direct, since the USSR reports only M surface-wave magnitudes determined from long-period horizontal instruments, whereas the NOS reports both body wave (mb) and, more recently, surface-wave magnitudes (Ms), also from long-period horizontal instruments but of a different type than the USSR seismographs.Two different approaches to the comparison have been made. For the first of these, the International Seismological Centre (ISC) Bulletin for 1964 was used as a data source, since it lists both NOS and USSR values for the reported events. Figs 1 to 4 show plots of magnitude residuals, USSR (M) values minus NOS (m,,) values, for various areas. Mean magnitude residuals and standard deviations * NOS Seismology staff transferred to Earth Sciences Laboratories of NOAA! Environmental